Ise Online Access For Introduction To Operations Research 11Th Edition Hillier Full Chapter

ISE EBook Online Access for
Introduction to Operations Research

11th Edition Hillier
Visit to download the full and correct content document:
https://ebookmass.com/product/ise-ebook-online-access-for-introduction-to-operation
s-research-11th-edition-hillier/
INTRODUCTION TO Page i
OPERATIONS RESEARCH
Page ii
INTRODUCTION TO Page iii
OPERATIONS RESEARCH
Eleventh Edition
FREDERICK S. HILLIER
Stanford University
GERALD J. LIEBERMAN
Late of Stanford University
Page iv
INTRODUCTION TO OPERATIONS RESEARCH

Published by McGraw-Hill Education, 2 Penn Plaza, New York, NY 10121. Copyright © 2021
by McGraw-Hill Education. All rights reserved. Printed in the United States of America. No part
of this publication may be reproduced or distributed in any form or by any means, or stored
in a database or retrieval system, without the prior written consent of McGraw-Hill Education,
including, but not limited to, in any network or other electronic storage or transmission, or
broadcast for distance learning.
Some ancillaries, including electronic and print components, may not be available to
customers outside the United States.
This book is printed on acid-free paper.
1 2 3 4 5 6 7 8 9 LWI 24 23 22 21 20
ISBN 978-1-260-57587-3
MHID 1-260-57587-X
Cover Image: Matt Diamond
All credits appearing on page or at the end of the book are considered to be an extension of
the copyright page.
The Internet addresses listed in the text were accurate at the time of publication. The
inclusion of a website does not indicate an endorsement by the authors or McGraw-Hill
Education, and McGraw-Hill Education does not guarantee the accuracy of the information
presented at these sites.
mheducation.com/highered
ABOUT THE AUTHORS Page v
Frederick S. Hillier was born and raised in Aberdeen, Washington, where he

was an award winner in statewide high school contests in essay writing,
mathematics, debate, and music. As an undergraduate at Stanford University, he
ranked first in his engineering class of over 300 students. He also won the
McKinsey Prize for technical writing, won the Outstanding Sophomore Debater
award, played in the Stanford Woodwind Quintet and Stanford Symphony
Orchestra, and won the Hamilton Award for combining excellence in engineering
with notable achievements in the humanities and social sciences. Upon his
graduation with a BS degree in industrial engineering, he was awarded three
national fellowships ( National Science Foundation, Tau Beta Pi, and Danforth) for
graduate study at Stanford with specialization in operations research. During his
three years of graduate study, he took numerous additional courses in
mathematics, statistics, and economics beyond what was required for his MS and
PhD degrees while also teaching two courses (including “Introduction to
Operations Research”). Upon receiving his PhD degree, he joined the faculty of
Stanford University and began work on the 1st edition of this textbook two years
later. He subsequently earned tenure at the age of 28 and the rank of full
professor at 32. He also received visiting appointments at Cornell University,
Carnegie-Mellon University, the Technical University of Denmark, the University of
Canterbury (New Zealand), and the University of Cambridge (England). After 35
years on the Stanford faculty, he took early retirement from his faculty
responsibilities in order to focus full time on textbook writing, and now is
Professor Emeritus of Operations Research at Stanford.
Dr. Hillier’s research has extended into a variety of areas, including integer
programming, queueing theory and its application, statistical quality control, the
application of operations research to the design of production systems, and
capital budgeting. He has published widely, and his seminal papers have been
selected for republication in books of selected readings at least 10 times. He was
the first-prize winner of a research contest on “Capital Budgeting of Interrelated
Projects” sponsored by The Institute of Management Sciences (TIMS) and the
U.S. Office of Naval Research. He and Dr. Lieberman also received the honorable
mention award for the 1995 Lanchester Prize (best English-language publication
of any kind in the field of operations research), which was awarded by the
Institute for Operations Research and the Management Sciences (INFORMS) for
the 6th edition of this book. In addition, he was the recipient of the prestigious
2004 INFORMS Expository Writing Award for the 8th edition of this book.
Dr. Hillier has held many leadership positions with the professional societies in
his field. For example, he has served as treasurer of the Operations Research
Society of America (ORSA), vice president for meetings of TIMS, co-general
chairman of the 1989 TIMS International Meeting in Osaka, Japan, chair of the
TIMS Publications Committee, chair of the ORSA Search Committee for Editor of
Operations Research, chair of the ORSA Resources Planning Committee, chair of
the ORSA/TIMS Combined Meetings Committee, and chair of the John von
Neumann Theory Prize Selection Committee for INFORMS. He also is a Fellow of
INFORMS. In addition, he served for 20 years (until 2013) as the founding series
editor for Springer’s International Series in Operations Research and Management
Science, a particularly prominent book series with nearly 300 published Page vi
books. In 2018, he was awarded the Kimball Medal (a lifetime
achievement award) by INFORMS for his distinguished contributions to the field
and to INFORMS.
In addition to Introduction to Operations Research and two companion
volumes, Introduction to Mathematical Programming (2nd ed., 1995) and
Introduction to Stochastic Models in Operations Research (1990), his books are
The Evaluation of Risky Interrelated Investments (North-Holland, 1969), Queueing
Tables and Graphs (Elsevier North-Holland, 1981, co-authored by O. S. Yu, with D.
M. Avis, L. D. Fossett, F. D. Lo, and M. I. Reiman), and Introduction to
Management Science: A Modeling and Case Studies Approach with Spreadsheets
(6th ed., McGraw-Hill, 2019, co-authored by his son Mark Hillier).
The late Gerald J. Lieberman sadly passed away in 1999. He had been
Professor E meritus of Operations Research and Statistics at Stanford University,
where he was the founding chair of the Department of Operations Research. He
was both an engineer (having received an undergraduate degree in mechanical
engineering from Cooper Union) and an operations research statistician (with an
AM from Columbia University in mathematical statistics, and a PhD from Stanford
University in statistics).
Dr. Lieberman was one of Stanford’s most eminent leaders. After chairing the
Department of Operations Research, he served as associate dean of the School of
Humanities and Sciences, vice provost and dean of research, vice provost and
dean of graduate studies, chair of the faculty senate, member of the University
Advisory Board, and chair of the Centennial Celebration Committee. He also
served as provost or acting provost under three different Stanford presidents.
Throughout these years of university leadership, he also remained active
professionally. His research was in the stochastic areas of operations research,
often at the interface of applied probability and statistics. He published
extensively in the areas of reliability and quality control, and in the modeling of
complex systems, including their optimal design, when resources are limited.
Highly respected as a senior statesman of the field of operations research, Dr.
Lieberman served in numerous leadership roles, including as the elected president
of The Institute of Management Sciences. His professional honors included being
elected to the National Academy of Engineering, receiving the Shewhart Medal of
the American Society for Quality Control, receiving the Cuthbertson Award for
exceptional service to Stanford University, and serving as a Fellow at Stanford’s
Center for Advanced Study in the Behavioral Sciences. In addition, the Institute
for Operations Research and the Management Sciences (INFORMS) awarded him
and Dr. Hillier the honorable mention award for the 1995 Lanchester Prize for the
6th edition of this book. In 1996, INFORMS also awarded him the prestigious
Kimball Medal for his distinguished contributions to the field and to INFORMS.
In addition to Introduction to Operations Research and two companion
volumes, Introduction to Mathematical Programming (2nd ed., 1995) and
Introduction to Stochastic Models in Operations Research (1990), his books are
Handbook of Industrial Statistics ( Prentice-Hall, 1955, co-authored by A. H.
Bowker), Tables of the Non-Central t-D istribution (Stanford University Press, 1957,
co-authored by G. J. Resnikoff), Tables of the Hypergeometric Probability
Distribution (Stanford University Press, 1961, co-authored by D. Owen),
Engineering Statistics , (2nd ed., Prentice-Hall, 1972, co-authored by A. H.
Bowker), and Introduction to Management Science: A Modeling and Case Studies
Approach with Spreadsheets (McGraw-Hill/Irwin, 2000, co-authored by F. S. Hillier
and M. S. Hillier).
ABOUT THE CASE Page vii
WRITERS
Karl Schmedders is professor of quantitative business administration at the

University of Zurich in Switzerland and a visiting associate professor at the
Kellogg Graduate School of Management (Northwestern University). His research
interests include management science, financial economics, and computational
economics and finance. He received his PhD in operations research from Stanford
University, where he taught both undergraduate and graduate classes in
operations research, including a case studies course in operations research. He
received several teaching awards at Stanford, including the university’s
prestigious Walter J. Gores Teaching Award. After post-doctoral research at the
Hoover Institution, a think tank on the Stanford campus, he became assistant
professor of managerial economics and decision sciences at the Kellogg School.
He was promoted to associate professor in 2001 and received tenure in 2005. In
2008, he joined the University of Zurich, where he currently teaches courses in
management science, business analytics, and computational economics and
finance. He has published research articles in international academic journals
such as Management Science, Operations Research, Econometrics, The Review of
Economic Studies, and The Journal of Finance, among others. He is a co-founder
of an EdTech Startup developing a digital learning and grading platform for science
education. At Kellogg he received several teaching awards, including the L. G.
Lavengood Professor of the Year Award. More recently he has won the best
professor award of the Kellogg School’s European EMBA program in eight different
years, as well as in 2017 for its EMBA program in Hong Kong.
Molly Stephens is a partner in the Los Angeles office of Quinn, Emanuel,
Urquhart & Sullivan, LLP. She graduated from Stanford University with a BS degree
in industrial engineering and an MS degree in operations research. Ms. Stephens
taught public speaking in Stanford’s School of Engineering and served as a
teaching assistant for a case studies course in operations research. As a teaching
assistant, she analyzed operations research problems encountered in the real
world and transformed these problems into classroom case studies. Her research
was rewarded when she won an undergraduate research grant from Stanford to
continue her work and was invited to speak at an INFORMS conference to present
her conclusions regarding successful classroom case studies. Following
graduation, Ms. Stephens worked at Andersen Consulting as a systems integrator,
experiencing real cases from the inside, before resuming her graduate studies to
earn a JD degree (with honors) from the University of Texas Law School at Austin.
She is a partner in the largest law firm in the United States devoted solely to
business litigation, where her practice focuses on complex financial and securities
litigation. She also is ranked as a leading securities litigator by Chambers USA
(2013 and 2014), which acknowledged “praise for her powerful and impressive
securities litigation practice” and noted that she is “phenomenally bright, a critical
thinker and great listener.”
DEDICATION Page viii
To the memory of our parents

and
To the memory of my beloved mentor,
Gerald J. Lieberman, who was one of the true
giants of our field
TABLE OF CONTENTS Page ix
PREFACE XX
CHAPTER 1
Introduction 1
1.1 The Origins of Operations Research 1
1.2 The Nature of Operations Research 3
1.3 The Relationship between Analytics and Operations Research 4
1.4 The Impact of Operations Research 8
1.5 Some Trends that Should Further Increase the Future Impact of Operations R
esearch 11
1.6 Algorithms and or Courseware 12
Selected References 14
Problems 14
CHAPTER 2
Overview of How Operations Research and Analytics Professionals Analyze Probl
ems 15
2.1 Defining the Problem 16
2.2 Gathering and Organizing Relevant Data 17
2.3 Using Descriptive Analytics to Analyze Big Data 18
2.4 Using Predictive Analytics to Analyze Big Data 19
2.5 Formulating a Mathematical Model to Begin Applying Prescriptive Analytics
23
2.6 Learning How to Derive Solutions from the Model 25
2.7 Testing the Model 28
2.8 Preparing to Apply the Model 29
2.9 Implementation 29
2.10 Conclusions 30
Problems 31
CHAPTER 3
Introduction to Linear Programming 32
3.1 Prototype Example 33
3.2 The Linear Programming Model 40
3.3 Assumptions of Linear Programming 45
3.4 Additional Examples 52
3.5 Formulating and Solving Linear Programming Models on a Spreadsheet 61
3.6 Formulating Very Large Linear Programming Models 69
3.7 Conclusions 77
Learning Aids for this Chapter on Our Website 77
Problems 78 Page x
Case 3.1 Reclaiming Solid Wastes 89
Previews of Added Cases on Our Website 89
Case 3.2Cutting Cafeteria Costs 89
Case 3.3Staffing a Call Center 89
Case 3.4Promoting a Breakfast Cereal 90
Case 3.5Auto Assembly 90
CHAPTER 4
Solving Linear Programming Problems: The Simplex Method 91
4.1 The Essence of the Simplex Method 92
4.2 Setting Up the Simplex Method 96
4.3 The Algebra of the Simplex Method 100
4.4 The Simplex Method in Tabular Form 106
4.5 Tie Breaking in the Simplex Method 111
4.6 Reformulating Nonstandard Models to Prepare for Applying the Simplex Met
hod 114
4.7 The Big M Method for Helping to Solve Reformulated Models 122
4.8 The Two-Phase Method is an Alternative to the Big M Method 129
4.9 Postoptimality Analysis 135
4.10 Computer Implementation 143
4.11 The Interior-Point Approach to Solving Linear Programming Problems 146
4.12 Conclusions 149
Appendix 4.1: An Introduction to Using LINDO and LINGO 149
Problems 154
Case 4.1 Fabrics and Fall Fashions 163
Case 4.2New Frontiers 165
Case 4.3Assigning Students to Schools 165
CHAPTER 5
The Theory of the Simplex Method 166
5.1 Foundations of the Simplex Method 166
5.2 The Simplex Method in Matrix Form 177
5.3 A Fundamental Insight 186
5.4 The Revised Simplex Method 189
5.5 Conclusions 192
Problems 193
CHAPTER 6
Duality Theory 200
6.1 The Essence of Duality Theory 200
6.2 Primal-Dual Relationships 208
6.3 Adapting to Other Primal Forms 213
6.4 The Role of Duality Theory in Sensitivity Analysis 217
6.5 Conclusions 220
Selected References 220 Page xi
Problems 221
CHAPTER 7
Linear Programming under Uncertainty 225
7.1 The Essence of Sensitivity Analysis 226
7.2 Applying Sensitivity Analysis 233
7.3 Performing Sensitivity Analysis on a Spreadsheet 250
7.4 Robust Optimization 259
7.5 Chance Constraints 263
7.6 Stochastic Programming with Recourse 266
7.7 Conclusions 271
Problems 273
Case 7.1 Controlling Air Pollution 281
Case 7.2Farm Management 282
Case 7.3Assigning Students to Schools, Revisited 282
Case 7.4Writing a Nontechnical Memo 282
CHAPTER 8
Other Algorithms for Linear Programming 283
8.1 The Dual Simplex Method 283
8.2 Parametric Linear Programming 287
8.3 The Upper Bound Technique 293
8.4 An Interior-Point Algorithm 295
8.5 Conclusions 306
Problems 307
CHAPTER 9
The Transportation and Assignment Problems 312
9.1 The Transportation Problem 313
9.2 A Streamlined Simplex Method for the Transportation Problem 326
9.3 The Assignment Problem 338
9.4 A Special Algorithm for the Assignment Problem 346
9.5 Conclusions 351
Problems 352
Case 9.1 Shipping Wood to Market 358
Case 9.2Continuation of the Texago Case Study 359
Case 9.3Project Pickings 359
Page xii
CHAPTER 10
Network Optimization Models 360
10.2 The Terminology of Networks 362
10.3 The Shortest-Path Problem 365
10.4 The Minimum Spanning Tree Problem 370
10.5 The Maximum Flow Problem 375
10.6 The Minimum Cost Flow Problem 383
10.7 The Network Simplex Method 391
10.8 A Network Model for Optimizing a Project’s Time-Cost Trade-Off 401
Problems 414
Case 10.1 Money in Motion 422
Case 10.2Aiding Allies 424
Case 10.3Steps to Success 424
CHAPTER 11
Dynamic Programming 425
11.1 A Prototype Example for Dynamic Programming 425
11.2 Characteristics of Dynamic Programming Problems 430
11.3 Deterministic Dynamic Programming 432
11.4 Probabilistic Dynamic Programming 448
Problems 455
CHAPTER 12
Integer Programming 460
12.2 Some BIP Applications 464
12.3 Using Binary Variables to Deal with Fixed Charges 470
12.4 A Binary Representation of General Integer Variables 472
12.5 Some Perspectives on Solving Integer Programming Problems 473
12.6 The Branch-and-Bound Technique and its Application to Binary Integer Progr
amming 477
12.7 A Branch-and-Bound Algorithm for Mixed Integer Programming 489
12.8 The Branch-and-Cut Approach to Solving BIP Problems 495
12.9 The Incorporation of Constraint Programming 502
12.10Conclusions 506
Problems 508
Case 12.1 Capacity Concerns 516
Previews of Added Cases on Our Website 518 Page xiii
Case 12.2Assigning Art 518
Case 12.3Stocking Sets 518
Case 12.4Assigning Students to Schools, Revisited Again 519
CHAPTER 13
Nonlinear Programming 520
13.1 Sample Applications 521
13.2 Graphical Illustration of Nonlinear Programming Problems 525
13.3 Types of Nonlinear Programming Problems 529
13.4 One-Variable Unconstrained Optimization 535
13.5 Multivariable Unconstrained Optimization 540
13.6 The Karush-Kuhn-Tucker (KKT) Conditions for Constrained Optimization 546
13.7 Quadratic Programming 550
13.8 Separable Programming 556
13.9 Convex Programming 563
13.10Nonconvex Programming (with Spreadsheets) 571
Problems 577
Case 13.1 Savvy Stock Selection 588
Case 13.2International Investments 589
Case 13.3Promoting a Breakfast Cereal, Revisited 589
CHAPTER 14
Metaheuristics 590
14.1 The Nature of Metaheuristics 591
14.2 Tabu Search 598
14.3 Simulated Annealing 608
14.4 Genetic Algorithms 618
Problems 630
CHAPTER 15
Game Theory 634
15.1 The Formulation of Two-Person, Zero-Sum Games 634
15.2 Solving Simple Games—A Prototype Example 636
15.3 Games with Mixed Strategies 641
15.4 Graphical Solution Procedure 643
15.5 Solving by Linear Programming 645
15.6 Extensions 649
Problems 651 Page xiv
CHAPTER 16
Decision Analysis 655
16.1 A Prototype Example 656
16.2 Decision Making without Experimentation 657
16.3 Decision Making with Experimentation 662
16.4 Decision Trees 668
16.5 Utility Theory 673
16.6 The Practical Application of Decision Analysis 680
16.7 Multiple Criteria Decision Analysis, Including Goal Programming 682
Problems 688
Case 16.1 Brainy Business 698
Preview of Added Cases on Our Website 700
Case 16.2Smart Steering Support 700
Case 16.3Who Wants to Be a Millionaire? 700
Case 16.4University Toys and the Engineering Professor Action Figures 700
CHAPTER 17
Queueing Theory 701
17.2 Basic Structure of Queueing Models 702
17.3 Some Common Types of Real Queueing Systems 707
17.4 The Role of the Exponential Distribution 708
17.5 The Birth-and-Death Process 714
17.6 Queueing Models Based on the Birth-and-Death Process 719
17.7 Queueing Models Involving Nonexponential Distributions 731
17.8 Priority-Discipline Queueing Models 739
17.9 Queueing Networks 744
17.10The Application of Queueing Theory 748
17.11Behavioral Queueing Theory 753
Problems 757
Case 17.1 Reducing In-Process Inventory 769
Preview of an Added Case on Our Website 770
Case 17.2Queueing Quandary 770
CHAPTER 18
Inventory Theory 771
18.1 Examples 772
18.2 Components of Inventory Models 774
18.3 Deterministic Continuous-Review Models 776
18.4 A Deterministic Periodic-Review Model 786
18.5 Deterministic Multiechelon Inventory Models for Supply Chain Management
791
18.6 A Stochastic Continuous-Review Model 810
18.7 A Stochastic Single-Period Model for Perishable Products 814 Page xv
18.8 Revenue Management 826
Problems 836
Case 18.1 Brushing Up on Inventory Control 846
Case 18.2TNT: Tackling Newsboy’s Teaching 848
Case 18.3Jettisoning Surplus Stock 848
CHAPTER 19
Markov Decision Processes 849
19.1 A Prototype Example 850
19.2 A Model for Markov Decision Processes 852
19.3 Linear Programming and Optimal Policies 855
19.4 Markov Decision Processes in Practice 859
Problems 863
CHAPTER 20
Simulation 866
20.1 The Essence of Simulation 866
20.2 Some Common Types of Applications of Simulation 878
20.3 Generation of Random Numbers 882
20.4 Generation of Random Observations from a Probability Distribution 886
20.5 Simulation Optimization 891
20.6 Outline of a Major Simulation Study 900
Problems 907
Case 20.1 Reducing In-Process Inventory, Revisted 912
Case 20.2Planning Planers 912
Case 20.3Pricing under Pressure 912
APPENDIXES
1. Documentation for the OR Courseware 913
2. Convexity 915
3. Classical Optimization Methods 920
4. Matrices and Matrix Operations 923
5. Table for a Normal Distribution 928
PARTIAL ANSWERS TO SELECTED PROBLEMS 930
INDEXES
Author Index 942
Subject Index 949
SUPPLEMENTS AVAILABLE Page ix
ON THE TEXT WEBSITE
www.mhhe.com/hillier11e
ADDITIONAL CASES
Case 3.2 Cutting Cafeteria Costs
Case 3.3 Staffing a Call Center
Case 3.4 Promoting a Breakfast Cereal
Case 3.5 Auto Assembly
Case 4.2 New Frontiers
Case 4.3 Assigning Students to Schools
Case 7.2 Farm Management
Case 7.3 Assigning Students to Schools, Revisited
Case 7.4 Writing a Nontechnical Memo
Case 9.2 Continuation of the Texago Case Study
Case 9.3 Project Pickings
Case 10.2 Aiding Allies
Case 10.3 Steps to Success
Case 12.2 Assigning Art
Case 12.3 Stocking Sets
Case 12.4 Assigning Students to Schools, Revisited Again
Case 13.2 International Investments
Case 13.3 Promoting a Breakfast Cereal, Revisited
Case 16.2 Smart Steering Support
Case 16.3 Who Wants to be a Millionaire?
Case 16.4 University Toys and the Engineering Professor Action Figures
Case 17.2 Queueing Quandary
Case 18.2 TNT: Tackling Newsboy’s Teaching
Case 18.3 Jettisoning Surplus Stock
Case 20.2 Planning Planers
Case 20.3Pricing under Pressure
SUPPLEMENT 1 TO CHAPTER 3
The LINGO Modeling Language
More about LINGO
SUPPLEMENT TO CHAPTER 6
An Economic Interpretation of the Dual Problem and the Simplex Method
Problem
A Case Study with Many Transportation Problems
The Construction of Initial BF Solutions for Transportation Problems
Problems
Some Innovative Uses of Binary Variables in Model Formulation
Problems
Preemptive Goal Programming and Its Solution Procedures
Problems
Case 16S-1 A Cure for Cuba
Case 16S-2 Airport Security
Stochastic Periodic-Review Models
Problems
A Policy Improvement Algorithm for Finding Optimal Policies
Problems
A Discounted Cost Criterion
Problems
Variance-Reducing Techniques
Problems
Regenerative Method of Statistical Analysis
Problems
CHAPTER 21
The Art of Modeling with Spreadsheets
21.1 A Case Study: The Everglade Golden Years Company Cash Flow Problem
21.2 Overview of the Process of Modeling with Spreadsheets
21.3 Some Guidelines for Building “Good” Spreadsheet Models
21.4 Debugging a Spreadsheet Model
21.5 Conclusions
Selected References
Learning Aids for this Chapter on Our Website
Problems
Case 21.1 Prudent Provisions for Pensions
CHAPTER 22
Project Management with PERT/CPM
22.1 A Prototype Example—The Reliable Construction Co. Project
22.2 Using a Network to Visually Display a Project
22.3 Scheduling a Project with PERT/CPM
22.4 Dealing with Uncertain Activity Durations
22.5 Considering Time-Cost Trade-Offs
22.6 Scheduling and Controlling Project Costs
22.7 An Evaluation of PERT/CPM
22.8 Conclusions
Selected References
Problems
Case 22.1 “School’s out forever . . .”
CHAPTER 23
Additional Special Types of Linear Programming Problems
23.1 The Transshipment Problem
23.2 Multidivisional Problems
23.3 The Decomposition Principle for Multidivisional Problems
23.4 Multitime Period Problems
23.5 Multidivisional Multitime Period Problems
23.6 Conclusions
Selected References
Problems
CHAPTER 24
Probability Theory
24.1 Sample Space
24.2 Random Variables
24.3 Probability and Probability Distributions
24.4 Conditional Probability and Independent Events
24.5 Discrete Probability Distributions
24.6 Continuous Probability Distributions
24.7 Expectation
24.8 Moments
24.9 Bivariate Probability Distribution
24.10 Marginal and Conditional Probability Distributions
24.11 Expectations for Bivariate Distributions
24.12 Independent Random Variables and Random Samples
24.13 Law of Large Numbers
24.14 Central Limit Theorem
24.15 Functions of Random Variables
Selected References
Problems
CHAPTER 25
Reliability
25.1 Structure Function of a System
25.2 System Reliability
25.3 Calculation of Exact System Reliability
25.4 Bounds on System Reliability
25.5 Bounds on Reliability Based upon Failure Times
25.6 Conclusions
Selected References
Problems
CHAPTER 26
The Application of Queueing Theory
26.1 Examples
26.2 Decision Making
26.3 Formulation of Waiting-Cost Functions
26.4 Decision Models
26.5 The Evaluation of Travel Time
26.6 Conclusions
Selected References
Problems
CHAPTER 27
Forecasting
27.1 Some Applications of Forecasting
27.2 Judgmental Forecasting Methods
27.3 Time Series
27.4 Forecasting Methods for a Constant-Level Model
27.5 Incorporating Seasonal Effects into Forecasting Methods
27.6 An Exponential Smoothing Method for a Linear Trend Model
27.7 Forecasting Errors
27.8 The ARIMA Method
27.9 Causal Forecasting with Linear Regression
27.10 Conclusions
Selected References
Problems
Case 27.1 Finagling the Forecasts
CHAPTER 28
Markov Chains
28.1 Stochastic Processes
28.2 Markov Chains
28.3 Chapman-Kolmogorov Equations
28.4 Classification of States of a Markov Chain
28.5 Long-Run Properties of Markov Chains
28.6 First Passage Times
28.7 Absorbing States
28.8 Continuous Time Markov Chains
Selected References
Problems
APPENDIX 6
Simultaneous Linear Equations
PREFACE Page xx
W hen Jerry Lieberman and I started working on the first edition

of this book, our goal was to develop a pathbreaking textbook
that would help establish the future direction of education in what
was then the emerging field of operations research. Following
publication, it was unclear how well this particular goal was met,
but what did become clear was that the demand for the book was
far larger than either of us had anticipated. Neither of us could have
imagined that this extensive worldwide demand would continue at
such a high level for such an extended period of time.
The enthusiastic response to our first ten editions has been most
gratifying. It was a particular pleasure to have the field’s leading
professional society, the international Institute for Operations
Research and the Management Sciences (INFORMS), award the 6th
edition honorable mention for the 1995 INFORMS Lanchester Prize
(the prize awarded for the year’s most outstanding English-
language publication of any kind in the field of operations research).
Then, just after the publication of the eighth edition, it was
especially gratifying to be the recipient of the prestigious 2004
INFORMS Expository Writing Award for this book, including receiving
the following citation:
Over 37 years, successive editions of this book have introduced more than
one-half million students to the field and have attracted many people to enter
the field for academic activity and professional practice. Many leaders in the
field and many current instructors first learned about the field via an edition
of this book. The extensive use of international student editions and
translations into 15 other languages has contributed to spreading the field
around the world. The book remains preeminent even after 37 years.
Although the eighth edition just appeared, the seventh edition had 46 percent
of the market for books of its kind, and it ranked second in international sales
among all McGraw-Hill publications in engineering.
Two features account for this success. First, the editions have been
outstanding from students’ points of view due to excellent motivation, clear
and intuitive explanations, good examples of professional practice, excellent
organization of material, very useful supporting software, and appropriate but
not excessive mathematics. Second, the editions have been attractive from
instructors’ points of view because they repeatedly infuse state-of-the-art
material with remarkable lucidity and plain language. For example, a
wonderful chapter on metaheuristics was created for the eighth edition.
When we began work on the first edition, Jerry already was a
prominent member of the field, a successful textbook writer, and
the chairman of a renowned operations research program at
Stanford University. I was a very young assistant professor just
starting my career. It was a wonderful opportunity for me to work
with and to learn from the master. I will be forever indebted to Jerry
for giving me this opportunity.
Now, sadly, Jerry is no longer with us. During the progressive
illness that led to his death in 1999, I resolved that I would pick up
the torch and devote myself to subsequent editions of this book,
maintaining a standard that would fully honor Jerry. Therefore, I
took early retirement from my faculty responsibilities at Stanford in
order to work full time on textbook writing for the foreseeable
Another random document with
no related content on Scribd:
“mere incidents.” From the review given below, in Chapter XIII, it is
clear that the main determinants of American culture accumulation,
after the first primitive start, were internal; and the case seems as
clear for metal working as for any phase.
109. Zero
One of the milestones of civilization is the number symbol zero.
This renders possible the unambiguous designation of numbers of
any size with a small stock of figures. It is the zero that enables the
symbol 1 to have the varying values of one, ten, hundred, or
thousand. In our arithmetical notation, the symbol itself and its
position both count: 1,234 and 4,321 have different values although
they contain the identical symbols. Such a system is impossible
without a sign for nothingness: 123 and 1,023 would be
indistinguishable. Our zero, along with the other nine digits, appears
to be an invention of the Hindus approximately twelve or fifteen
hundred years ago. We call the notation “Arabic” because it was
transmitted from India to Europe by the Arabs.
Fig. 28. Maya symbols for zero: a, monumental; b, c, cursive. (From Bowditch.)
Without a zero sign and position values, two methods are open for
the representation of higher numerical values. More and more signs
can be added for the high values. This was done by the Greeks and
Romans. MV means 1,005, and only that. This is simple enough; but
1,888 requires so cumbersome a denotation as MDCCCLXXXVIII—
thirteen figures of six different kinds. A simple system of multiplying
numbers expressed like this one is impossible. The unwieldiness is
due to the fact that the Romans, not having hit upon the device of
representing nothingness, employed the separate signs I, X, C, M for
the quantities which we represent by the single symbol 1 with from
no to three zeroes added.
The other method is that followed by the Chinese. Besides signs
corresponding to our digits from 1 to 9, they developed symbols
corresponding to “ten times,” “hundred times,” and so on. This was
much as if we should use the asterisk, *, to denote tens, the dagger,
† , for hundreds, the paragraph, ¶, for thousands. We could then
represent 1,888 by 1 ¶ 8 † 8 * 8, and 1,005 by 1 ¶ 5, without any risk
of being misunderstood. But the writing of the numbers would in
most cases require more figures, and mathematical operations
would be more awkward.
The only nation besides the Hindus to invent a zero sign and the
representation of number values by position of the basic symbols,
were the Mayas of Yucatan. Some forms of their zero are shown in
Figure 28. This Maya development constitutes an indubitable parallel
with the Hindu one. So far as the involved logical principle is
concerned, the two inventions are identical. But again the concrete
expressions of the principle are dissimilar. The Maya zero does not
in the least have the form of our or the Hindus’ zero. Also, the Maya
notation was vigesimal where ours is decimal. They worked with
twenty fundamental digits instead of ten. Their “100” therefore stood
for 400, their “1,000” for 8,000.[17] Accordingly, when they wrote, in
their corresponding digits, 1,234, the value was not 1,234 but 8,864.
Obviously there can be no question of a common origin for such a
system and ours. They share an idea or a method, nothing more. As
a matter of fact, these two notational systems, like all others, were
preceded by numeral word counts. Our decimal word count is based
on operations with the fingers, that of the Maya on operations with
the fingers and toes. Twenty became their first higher unit because
twenty finished a person.
It is interesting that of the two inventions of zero, the Maya one
was the earlier. The arithmetical and calendrical system of which it
formed part was developed and in use by the time of the birth of
Christ. It may be older; it certainly required time to develop. The
Hindus may have possessed the prototypes of our numerals as early
as the second century after Christ, but as yet without the zero, which
was added during the sixth or according to some authorities not until
the ninth century. This priority of the Maya must weaken the
arguments sometimes advanced that the ancient Americans derived
their religion, zodiac, art, or writing from Asia. If the zero was their
own product, why not the remainder of their progress also? The only
recourse left the naïve migrationist would be to turn the tables and
explain Egyptian and Babylonian civilization as due to a Maya
invasion from Yucatan.
110. Exogamic Institutions

In many parts of the world nations live under institutions by which
they are divided into hereditary social units that are exogamous to
one another. That is, all persons born in a unit must take spouses
born in some other unit, fellow members of one’s unit being regarded
as kinsmen. The units are generally described as clans, gentes, or
sibs; or, where there are only two, as moieties. In many cases the
sibs or moieties are totemic; named after, or in some way associated
with, an animal, plant, or other distinctive object that serves as a
badge or symbol of the group. Often the association finds expression
in magic or myth. Since under this system one is born into his social
unit, cannot change it, and can belong to one only, it follows that
descent is unilateral. It is impossible for a man to be a member of
both his father’s and his mother’s sib or totem; custom has
established everywhere a rigid choice between them. Some tribes
follow descent from the mother or matrilinear reckoning, others are
patrilinear.[18]
Institutions of this type have a wide and irregular distribution. They
are frequent in Australia, New Guinea, and Melanesia; found in parts
of the East Indies and southeastern Asia; quite rare or stunted in the
remainder of Asia and Polynesia; fairly common in Africa, though
they occur in scattered areas; characteristic again of a large part of
North America, but confined to a few districts of South America. At a
rough guess, it might be said that about as many savage peoples,
the world over, possess totemic-exogamous clans or moieties as
lack them. The patchiness on the map of exogamic institutions
argues against their being all the result of a wave of culture
transmission emanating from a single source. Had such a diffusion
occurred, it should have left its marks among the numerous
intervening tribes that are sibless. Further, both in the eastern and
western hemispheres, the most primitive and backward tribes are,
with fair regularity, sibless and non-totemic. If therefore a
hypothetical totem-sib movement had encircled the planet, it could
not have been at an extremely ancient date, else the primitive tribes
would have been affected by it; and since records go back five
thousand years in parts of the Mediterranean area, the movement, if
relatively late, should have left some echo in history, which it has not.
Fig. 29. Distribution of types of exogamic institutions in Australia: 2M, two classes,
matrilinear; 4M, four classes, matrilinear; 4P, four classes, patrilinear; 8P,
eight classes, patrilinear; black areas, no classes, patrilinear exogamic
totems; X, totems independent of classes; Y, totems replace sub-classes; Z,
no organization; ?, uninhabited or unknown. (After Thomas and Graebner.)
It is therefore probable that totem-sib institutions did not all

emanate from one origin, but developed independently several
times. The question then becomes, how often, and where?
The evidence for America has been reviewed in another
connection (§ 185). It can be summarized in the statement that at
least two of the three sib areas[19] of North America, and probably
the two principal ones of South America, seem to have resulted from
a single culture growth which perhaps centered at one time, although
subsequently superseded, in the middle sector of the double
continent. This movement may have had first a patrilinear and then a
matrilinear phase, though at no great interval of time. The third North
American area may have got its patrilinear sib institutions from the
same source but probably developed its matrilinear ones locally as a
subsequent growth. If so, this would be an instance of convergence
on the same continent—a rather rare phenomenon.
For Australia, New Guinea, and Melanesia, the geographical
proximity is so close as to suggest a single origin for the whole area.
Patrilinear and matrilinear descent are both found in Australia as well
as Melanesia. This fact has been interpreted as the result of an
earlier patrilineal and a later matrilineal phase of diffusion. It is
interesting that this conclusion parallels the tentative one
independently arrived at for America, although in both hemispheres
further analysis and distributional study must precede a positive
verdict.
In the principal other sib area, Africa, the reckoning is so
prevailingly patrilineal, that the few cases of matrilineate can
scarcely be looked upon as anything but secondary local
modifications. As to whether the totemism and exogamy of Africa
can be genetically connected with those of Australia-Melanesia, it is
difficult to decide. The more conservative attitude would be to regard
them as separate growths, although so many cultural similarities
have been noted between western Africa and the area that stretches
from Indo-China to Melanesia, as to have raised suspicions of an
actual connection (§ 270). Yet even if these indications were to be
confirmed, thus sweeping most or all the Old World sib institutions
into a single civilizational movement, the distinctness of this from the
parallel development of the New World would remain.
It is significant that in the three successive continents of America,
Oceania, and Africa the patrilinear and matrilinear phases of the sib
type of society exist side by side, and that the same duality even
holds for each of the separate areas in America. That is, the
Northwest American sib area includes matrilinear as well as
patrilinear tribes; the Southwest area includes both; and so on.
A similar tendency toward geographical association is found in
other phases of social structure: the clan and moiety, and again
totemism and exogamy.
The clan or multiple form of sib organization is logically distinct
from the moiety or dual form. Under the plural system, a person,
being of clan A, may marry at will into clans B, C, D, E, F. Three of
his four grandparents would normally be of other clans than his own,
but of which they were members, would vary in each individual case.
In a patrilineal society, one member of clan A would have his
maternal uncles of clan B; the next, of clan C; a third, perhaps of
clan F; according to the choices which their fathers had made of
wives.
Under the dual system, however, a member of moiety A may just
as well be regarded as having a wife of moiety B prescribed or
predestined for him as being forbidden an A wife. Two of his
grandparents, say his father’s father and his mother’s mother, are
inevitably of his own moiety, the two others of the opposite one.
Every possible kinsman—his maternal uncle, his cross cousin, his
father-in-law, his wife’s brother-in-law, his daughter’s son—has his
moiety affiliation foreordained. Where descent is paternal, for
instance, everybody knows that his future mother-in-law must be of
his own moiety. Evidently the effect of this dual system on the
relations between kinsfolk, on social usages, on the individual’s
attitude of mind toward other individuals, should normally tend to be
profoundly different from the influence of a multiple clan system. On
theoretical grounds it might seem likely that the dual and multiple
schemes had nothing to do with each other, that they sprang from
distinct psychological impulses.
Yet such a belief would be ungrounded, as the facts of distribution
promptly make clear. In every multiple sib area of any moment,
moieties also occur, and vice versa. In the California-Southwest
region, for instance, tribes like the Miwok are divided into moieties
only, the Mohave and Hopi into clans only, the Tewa and Cahuilla
into moieties subdivided into clans. So in the Eastern, the Plains,
and the Northwest areas of North America, clan tribes and moiety
tribes live side by side; whereas as soon as these regions are left
behind, there are vast districts—much of Mexico, Texas, the Great
Basin and Plateau, northern Canada and the Arctic coast—whose
inhabitants get along without either clans or moieties. So again in
Melanesia and in Australia (Fig. 29), the two types of organization
exist side by side, while most of Polynesia, Asia, and Europe are
void of both. Only Africa shows some development of multiple clan
institutions but no moieties. In short, as soon as areas of some size
are considered, they prove in the main to be of two kinds. Either they
contain both clan tribes and moiety tribes, or they contain neither.
That is, the clan institution and the moiety institution are correlated or
associated in geography, as patrilinear and matrilinear descent are
correlated, which indicates a community of origin for them.
A similar relation exists between exogamic units, be they moieties
or clans, and totemism. The first constitutes a scheme of society, a
method of organization; the second, a system of symbolism. Sibs are
social facts, totems a naming device with magico-religious
implications. There is no positive reason why they should be
associated. They are not always associated. There are American
tribes like the Navaho and Gros Ventre that live under unilateral and
exogamic institutions without totems. Placenames or nicknames
distinguish the groups. In Australia, the Arunta possess unilaterally
reckoning exogamic groups as well as totems, but the two are
dissociated; a person takes his group by descent, his totem wholly
irrespective of this according to place of birth or conception. In Africa
there are no less than six tribes or series of tribes in which exogamy
and totemism are thus dissociated; a person takes his totem from his
father, his exogamic unit from his mother, so that the two ordinarily
do not coincide for parent and child. Exogamy and totemism, then,
are theoretically separate factors.
Yet since they are distinct, it is remarkable that in probably seven
or eight tenths of all cases they coincide, and that in each of the
continents or areas containing them they are found associated. If
exogamy and totemism had grown out of separate roots, one could
expect at least one considerable area somewhere in which one of
them appeared without the other. But there is no such area.
Wherever social exogamy appears among a larger group of nations,
social totemism also crops out; and vice versa.
It must then be concluded that exogamy and totemism,
matrilineate and patrilineate, multiple and dual sibs, all show a strong
tendency toward association with one another. In other words, their
correlation is positive and strong. Even where they seem mutually
exclusive in their very nature, like matrilinear and patrilinear
reckoning, ways have been found by unconscious human ingenuity
to make them coexist among one people, as when one reckoning is
attached to the exogamy, the other to the totemism; and still more
often they occur among adjacent tribes.
111. Parallels and Psychology

Such associations as these are common enough in the history of
civilization. A number are touched upon elsewhere in this volume
under the name of culture trait associations or complexes (§ 97,
149). But usually such a complex or nexus consists of culture
elements that have no necessary connection: Christianity and
trousers, for instance. It is accident that first throws them together;
association ties them one to the other; once the cluster is
established by usage, its coherence tends to persist. But there is
something arbitrary about this cohesion, generally. There is no
inherent reason why a hundred American tribes that grow maize
should also grow beans and squashes and nothing else; but they do
limit themselves to the three. The distinctive feature of the sib-
complex is that it has an almost reasonable quality. Its elements,
however separate or even opposite logically, do have a certain
psychological affinity to one another. Also, the arbitrary maize-
beans-squash complex and other complexes are generally not
duplicated. But the intricate and psychologically founded totemism-
exogamy-descent complex looks as if it might have been triplicated
or quadruplicated. This parallelism, if the facts prove to substantiate
it, is parallelism raised to a higher power than any yet considered.
Heretofore the discussion has been of the parallelism of single
culture traits. Here it is a case of parallelism of a complex of culture
traits. Such complex convergence might suggest something peculiar
to or inherent in the human mind, leading it, once it is stimulated to
commence the development of one of the factors of the complex, to
follow with the production of the other factors.[20]
Similar instances would be the tendency of agriculture to be
followed by town life, if it could be demonstrated, though this seems
doubtful; of settled living to be accompanied by migration legends; of
religions with personal founders to become propagandizing and
international but in time to die out among the nations in which they
were originated.
In regard to all such cases it may be said first of all that an
exhaustive analysis is necessary to ascertain whether the seeming
association or correlation is borne out by the facts. Second, the
possibility of diffusion must be eliminated. If Melanesian and African
totem-exogamy are both products of one culture growth, they cannot
be counted as two examples of the same association. If they should
ultimately both prove to be linked with the American system by a
wave of migration or culture contact, as has, indeed been maintained
in two separate hypotheses recently advanced, parallelism is of
course disproved altogether. But such views are as yet
undemonstrated and seem extreme; and if, after continued search of
the evidence, two or more such associations or complex parallels as
the exogamic-totemic scheme of society stand as independent
growths, it is evident that they will be something in the nature of
cultural manifestations of psychological forces. In short, we should
then be beginning to grasp specific psychological determinants for
the phenomena or events of civilization. But as yet such a causal
explanation of the data of anthropology by the mechanism of
psychology has not been achieved.
112. Limitations on the Principle

From the evidences reviewed in this and the last chapter, the
conclusion is confirmed which social philosophers had long since
reached, that imitation is the normal process by which men live, and
that invention is rare, a thing which societies and individuals oppose
with more resistance than they are ever aware of, and which
probably occurs only as the result of the pressure of special
circumstances, although these are as yet little understood. Not only
are a hundred instances of diffusion historically traceable for every
one of parallelism, but the latter is regularly limited in scope.
Something tends to make us see phenomena more parallel than
they actually are. They merely spring from the same impulse, they
inhere in the properties of objects or nature, they bear resemblance
at one point only—and differ at all other points. Yet they tend to
impress us, in some mysterious way, as almost identical. The history
of civilization has no more produced two like cultures, or two
separately developed identical culture traits, than has the evolution
of organic life ever duplicated a species by convergently modifying
two distinct forms. A whale may look fishlike, he is a mammal. The
Hindu and the Maya zero are logically the same; actually they have
in common nothing but their abstract value: their shapes, their place
in their systems, are different. The most frequent process of culture
history therefore is one of tradition or diffusion in time and space,
corresponding roughly to hereditary transmission in the field of
organic life. Inventions may be thought of as similar to organic
mutations, those “spontaneous” variations that from time to time
arise and establish themselves. The particular causes of both
inventions and mutations remain as good as unknown. Now and
then a mutant or an invention heads in the same direction as another
previously arisen one. But, since they spring from different
antecedents, such convergences never attain identity. They remain
on the level of analogous resemblance. Substantial identity, a part
for part correspondence, is invariably a sign of common origin, in
cultural as well as organic history.
CHAPTER X
THE ARCH AND THE WEEK
113. House building and architecture.—114. The problem of spanning.—

115. The column and beam.—116. The corbelled arch.—117. The true
arch.—118. Babylonian and Etruscan beginnings.—119. The Roman
arch and dome.—120. Mediæval cathedrals.—121. The Arabs: India:
modern architecture.—122. The week: holy numbers.—123.
Babylonian discovery of the planets.—124. Greek and Egyptian
contributions: the astrological combination.—125. The names of the
days and the Sabbath.—126. The week in Christianity, Islam, and
eastern Asia.—127. Summary of the diffusion.—128. Month-thirds
and market weeks.—129. Leap days as parallels.
In exemplification of the principles discussed in the last two

chapters, the next two are given over to a more detailed
consideration of several typical ramifying growths whose history
happens to be known with satisfactory fullness. These are the arch,
the week, and the alphabet.
113. House Building and Architecture

The history of human building makes a first impression of an
endless tangle. Every people rears some sort of habitations, and
however rude these are, structural principles are involved. Obviously,
too, geography and climate are bound to have at least a delimiting
influence. The Eskimo of the Arctic cannot build houses of wood; the
inhabitants of a coral reef in the Pacific could not, however much
they might wish, develop a style in brick. In structures not used as
dwellings, their purpose necessarily affects their form. A temple is
likely to be made on a different plan from a court of law. Temples
themselves may vary according to the motives and rituals of the
religions which they serve.
Bewilderment begins to abate as soon as one ceases trying to
contemplate all buildings reared by human hands. Obviously a
dwelling erected by a small family group for the utilitarian purpose of
shelter is likely to be more subject to immediate adaptations to
climate than a large communal structure serving some purpose such
as the service of a deity. If consideration be restricted still further, to
religious or public buildings set up with the idea of permanence,
another class of causes making for variability begins to be
eliminated. A structure intended as an enduring monument is reared
with consideration to the impression that it will create in the minds of
future generations. Its emotional potentialities, be these evoked by
its mere size, by the æsthetic nature of its design, or by a
combination of the two, come into the forefront. Such permanent
buildings being in stone or brick, techniques which flourish in wood
or other temporary materials are eliminated. Finally, a monumental
structure is possible only at the hands of a community of some size.
An unstable group of nomads, a thinly scattered agricultural
population, cannot assemble in sufficient numbers even for periods
each year, to carry out the long-continued labors that are necessary.
The aggregation of numbers of men in one spot is always
accompanied by specialization in advancement of the arts.
Consequently the very fact that a structure is monumental involves
the probability that its builders are able to rise above the limitations
of mere necessity, and can in some degree execute products of their
imagination.
114. The Problem of Spanning

If now our attention be confined to large buildings of the more
massive and permanent sort, it becomes clear that one of the chief
problems which all their constructors have had to grapple with, is
that of roofing large spaces and spanning wide openings in walls. A
pyramid can be heaped up, or a wall reared to a great height, without
much other than quantitative difficulties being encountered. A four
hundred foot pyramid does not differ in principle from the waist-high
one that a child might pile up. The problems which it involves are
essentially the economic and political ones of providing and
controlling the needed multitudes of workers. Architecture as such is
in abeyance and the engineering problems involved are mainly those
of transporting and raising large blocks of stone. Much the same
holds of walls. The Incas, for instance, reared masonry of astounding
massiveness and exactness without ever seriously attempting to
solve architectural problems.
Once, however, a structure is planned to cover a wide space, it
becomes architecturally ambitious. The roof of a large dwelling can
be made easily of poles and thatch by such collaborators as a family
might muster. But to span a clear space of some size in stone
requires more than numbers of workers. The accomplishment also
yields definite sense of achievement which is strong in proportion as
the extent of the ceiling is great. The difficulties are diminished in
proportion as the mass of the structure is large and the clear space
is small, but the satisfying effect is correspondingly decreased. A
vault whose walls are thicker than its interior is wide, produces as
chief impression an effect of massiveness. One feels the solidity of
the structure, the amount of labor that has gone into it; but one is left
without the sense of a worth-while difficulty having been self-
imposed and mastered. Sooner or later, therefore, after men began
to hold themselves available for co-operative enterprises in numbers,
adventurous minds must have been fired with a desire to grapple
with problems of æsthetic construction, and to leave behind them
monuments of triumphant solution. The story of these voluntary and
imaginative endeavors is the history of monumental art.
Two principal methods have been followed in the solution of the
problem of covering large free spaces. The first is the method of the
column and the lintel; the second that of the arch or vault. The
column and lintel do not differ fundamentally from the idea of the wall
with superimposed roof beams. The elements of both are vertical
support and horizontal beam. In the arch, however, this simple
scheme is departed from, and the covering elements take on a
curved or sloping form. The apparently free float of the span is
stimulatingly impressive, especially when executed in a heavy and
thoroughly rigid material. The beam is subject to bending stress.
Timber makes a good material because of its strength against
breakage by bending. Stone is unreliable or outrightly weak against
a bending stress, besides adding to the stress by its own weight.
There are therefore inherent limitations on the space that can be
covered by a horizontal stone beam.
115. The Column and Beam

Most early architecture developed the column. Even so superb an
architecture as that of the Greeks never rose above it. The æsthetic
value of the Parthenon lies in the balance and feeling with which a
fundamentally simple plan has been elaborated, not in the daring
way in which an inherently ambitious problem has been met.
On account of its essential simplicity, columnar architecture grew
up among several historically unconnected nations. In the case of
most of them, there can be distinguished an early stage of building in
wood, when the column was the trunk of a tree, and a later stage in
which the post was replaced by a monolith, or by superimposed
drums of stone. This change appears to have taken place somewhat
independently in Egypt and in Greece, and wholly so in Mexico. It
has been thought that Greek architecture was derived from Egypt,
but there was probably little more than a transmission of stimulus,
since Greek temples were wooden pillared several thousand years
after the Egyptians were rearing huge stone columns. Furthermore, if
the Greeks had borrowed their column outright from Egypt, they
would probably have copied it slavishly at the outset. Yet their early
capitals are without the lotus flower head in which the Egyptian
column terminated. Here, then, and still more in Mexico, there was
parallel development.
The failure of the Greeks to pass beyond column and lintel
architecture may seem strange for a people that showed so unusual
an artistic faculty and so bold and enterprising a spirit as they
manifested in most departments of civilization. The cause appears to
lie not in any internal arrest of their artistic evolution, but in the
conditions that prevailed in another field of their culture: their political
particularity. The Greek state remained a city. All attempts to
establish larger political aggregates, whether on the basis of
confederation or conquest, failed miserably and speedily. The Greek
was ingrainedly addicted to an outlook that was not merely provincial
but literally municipal. The result was that really large coöperative
enterprises were beyond him. Paved roads, aqueducts, sewers, and
works of a like character were scarcely attempted on any scale of
magnitude. With the rather small numbers of individuals which at
best the Greeks assembled in one spot, such works were not
necessary, and undertaken in mere ambition, they would have
encountered public antagonism. Consequently Greek public
buildings were, by the standards of many other nations, mediocre in
size of ground plan, low in height, without endeavor to impress by
sweep of clear space. This fact illustrates the almost organic
interconnection existing between the several sides of the culture of
any people; it illustrates also the importance of knowing the whole of
a civilization before trying to provide an explanation for any one of its
manifestations.
116. The Corbelled Arch

The arch brings in an inherently new principle of architecture. It is
a device for carrying construction over an empty space without
horizontal beams. But it may take two principal forms: the corbelled
or “false” arch, and the “true” arch. Both are arches in form, but the
blocks that form the curvature of one are not self-supporting; in the
other they are.
The corbelled arch achieves its span through a successive
projection of the stones or bricks that abut on each side of the open
space. The stone at the end of the second course of masonry
extends part of its length beyond the end stone of the first course. At
the opposite side, the second course hangs similarly out above the
first. In the third course, the end blocks again project beyond those
of the second. The arrangement thus is that of two series of
brackets, or two staircases turned upside down. The higher the
masonry rises, the more do the clear space narrow and the two lines
of hanging steps approach until they meet and the arch is complete.
What keeps the projecting stones from toppling into the clear space?
Nothing, obviously, but such weight as is put on their inner or
embedded ends. Suppose a stone projects a third of its length
beyond the one below, so that its center of gravity is still above the
lower stone. It will then lie as placed. Suppose still another stone
again projects a third of its length beyond the second. Its center of
gravity now falling outside the lowest block, it will topple both itself
and the second one. Only if other blocks are inserted behind will
their counterweight hold up the projecting blocks. Obviously, there
will be more such counterweights needed the higher the side of the
arch rises. In general, the area of wall needed as counterweight is at
least as great as the area of overhanging. If the arch is to clear ten
feet horizontally—hanging over five feet from each side—there must
be five feet or more of masonry built up on each side of the clear
space. A corbelled arch forming a relatively small doorway in the
face of a wall presents no difficulty, but a corbelled arch that stands
free is impossible.
The same principle holds for the vault, which is a three-
dimensional extension of the virtually two-dimensional arch. The
hollow or half-barrel of the corbelled vault has to be flanked by a
volume of building material exceeding its own content. This need
eliminates corbelling as a possible method of rearing structures that
rise free and with lightness. Hence the clumsy massiveness of, for
instance, Maya architecture, which, so far as it employs the vault,
often contains more building material than spanned space.
Another difficulty, beyond that of counterweighting, which besets
the user of the corbelled arch, is that the projecting stones of each
course are subjected to the same bending strain as a beam. The
weight above strives to snap them in two.
The corbelled arch and vault have been independently devised
and have also diffused. They were employed in gigantic Bronze age
tombs at Mycenæ in Greece—the so-called treasure house of
Atreus,—in Portugal, and in Ireland (Fig. 41). These developments
seem historically connected. On the other hand the Mayas of
Yucatan also built corbelled arches, which must constitute a
separate invention. This parallel development differs from that of the
true arch, which seems everywhere to be derived from a single
original source.
117. The True Arch

The true arch differs from the corbelled in needing no
counterweight. The blocks that form the under surface or soffit of its
span are self-sustaining. The true arch thus yields an æsthetic
satisfaction which can be attained in no other way, especially when it
soars in magnitude. The fundamental principle of the true arch is the
integration of its elements. Such an arch is nothing until completed;
but from that moment its constituents fuse their strength. Each block
has a shape which is predetermined by the design of the whole, and
each is useless, in fact, not even self-supporting, until all the others
have been fitted with it. Hence the figure of speech as well as the
reality of the keystone: the last block slipped into place, locking itself
and all the others. The features of the blocks or “voussoirs” which
makes possible this integration, is the taper of their sides. Each is a
gently sloping piece of wedge instead of a rectangular block. When
bricks replace dressed stone, the mortar takes the place of this
shaping, being thinner toward the inner face of the vault and thicker
toward the interior of the construction.
A true arch in process of erection would instantly collapse if not
held up. It can be built only over a scaffold or “centering.” Once
however the keystone has wedged its parts together, it not only
stands by itself but will support an enormous weight. The greater the
pressure from above, the more tightly are the blocks forced together.
Instability in a true arch is not due to the bending stress coming from
the superimposed mass, as in the corbelled arch or a horizontal
roofing. The blocks are subjected only to crushing pressure, which
stone and brick are specially adapted to withstand. The weakness of
the arch is that it turns vertical into horizontal thrust. With more
weight piled on top, the sidewise thrust, the inclination to spread
apart, becomes greater, and must be resisted by buttressing. This is
what the Hindus mean when they say that “the arch never sleeps.”
118. Babylonian and Etruscan Beginnings
While the exact circumstances attending the invention of the true
arch are not clear, the earliest specimens preserved are from the
ancient brick-building peoples of Babylonia, especially at Nippur
about 3,000 B.C. Thence the principle of the arch was carried to
adjacent Assyria. Both these Mesopotamian peoples employed the
arch chiefly on a small scale in roofing doors and in tunnels. It
remained humble and utilitarian in their hands; its architectural
possibilities were scarcely conceived. They continued to rear their
monumental structures mainly with an eye to quantity: high and thick
walls, ramps, towers ascending vertically or by steps, prevailed.
The true arch and vault are next found in Italy, among a
prosperous city-dwelling people, the Etruscans, some seven or more
centuries before Christ. All through the civilization of this nation runs
a trait of successful but never really distinctive accomplishment. The
Etruscans were receptive to new ideas and applied them with
energy, usually only to degenerate them in the end. Whether they
discovered the arch for themselves or whether knowledge of it was
carried to Italy from Asia is not wholly clear, since history knows little
about the Etruscans, and archæology, though yielding numerous
remains, leaves the problem of their origin dark. The Etruscans, or
Tyrrhenians as the Greeks knew them, were however active traders,
and a number of features in their civilization, such as liver divination
(§ 97), as well as ancient tradition, connect them with Asia. It is
therefore probable that the principle of arch construction was
transmitted to them from its earlier Babylonian source. The
Etruscans also failed to carry the use of the arch far into monumental
architecture. They employed it in tombs, gates, and drains rather
than as a conspicuous feature of public buildings.
119. The Roman Arch and Dome

From the Etruscans their neighbors, the Romans, learned the
arch. They too adopted it at first for utilitarian purposes. The great
sewer of Rome, for instance, the Cloaca Maxima, is an arched vault
of brick. Gradually, however, as the Romans grew in numbers and
wealth and acquired a taste for public undertakings, they transferred
the construction to stone and introduced it into their buildings. By the
time their polity changed from the republican to the imperial form, the
arch was the most characteristic feature of their architecture. The
Greeks had built porticos of columns; the Romans erected frontages
of rows of arches. The exterior of their circus, the Coliseum, is a
series of stories of arches. Much of the mass of the structure also
rests upon arches, thus making possible the building of the huge
edifice with a minimum of material. On the practical side, this is one
of the chief values of the arch. The skill which evolved it eliminates a
large percentage of brute labor. Earlier peoples would have felt it
necessary to fill the space between the interior tiers of seats and the
outer wall of the Coliseum.
Once the fever of architecture had infected them, the Romans
went beyond the simple arch and vault. They invented the dome. As
the simplest arch, such as a doorway or window, a perforation in a
wall, is essentially two dimensional, and a vault is the projecting of
this plane area into the three dimensions of a half cylinder, so the
dome can be conceived as the extension of the arch into another
three-dimensional form, the half sphere. Their relations are those of
a hoop, a barrel, and a hollow ball. Imagine a vault revolved on a
central vertical pivot, and it will describe the surface of a dome. Two
intersecting arches can be served by a single keystone.
Theoretically, more and more arches can be introduced to intersect
at the same point, until they form a continuous spheroid surface.
Neither construction nor the evolution of the dome did actually take
place by this method of compounding arches, which however serves
to illustrate the logical relation of the two structures.
The Roman engineers put domes on their Pantheon, the tomb of
Hadrian, and other buildings. In the centuries in which the
Mediterranean countries were Romanized, the dome and the arch,
the vault and the row of arches set on pillars, became familiar to all
the inhabitants of the civilized western world. After Roman power
crumbled, the architectural traditions survived. Even when there was

Ise Online Access For Introduction To Operations Research 11Th Edition Hillier Full Chapter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ise Online Access For Introduction To Operations Research 11Th Edition Hillier Full Chapter

Uploaded by

Copyright:

Available Formats

ISE EBook Online Access for

Introduction to Operations Research

INTRODUCTION TO OPERATIONS RESEARCH

Frederick S. Hillier was born and raised in Aberdeen, Washington, where he

Karl Schmedders is professor of quantitative business administration at the

To the memory of our parents

PARTIAL ANSWERS TO SELECTED PROBLEMS 930

W hen Jerry Lieberman and I started working on the first edition

110. Exogamic Institutions

It is therefore probable that totem-sib institutions did not all

111. Parallels and Psychology

112. Limitations on the Principle

113. House building and architecture.—114. The problem of spanning.—

In exemplification of the principles discussed in the last two

113. House Building and Architecture

114. The Problem of Spanning

115. The Column and Beam

116. The Corbelled Arch

117. The True Arch

119. The Roman Arch and Dome

You might also like